Signaling Michigan’s full commitment to expanding resources for data scientists and attracting top talent in the field, the Michigan Institute for Data Science (MIDAS) at UofM holds annual symposiums that bring together some of the best minds in the field from all over the world. After several successful years, with a growing list of attendees coming to mine the jewels of data science insight coming from the growing list of esteemed keynote speakers, 2020 will see the first ever Women+ Data Science Symposium hosted right at the UofM campus in Ann Arbor.<!- mfunc feat_school ->
It’s all driven by a huge expansion in the field overall as demand spikes for data-driven solutions in global industry and government, and the skilled data scientists capable of turning all that raw data into gold.
The skills gap between what those organizations need and the number of data scientists in the job market today is immense. According to tech industry recruiting firm, DICE, data engineer and senior data scientist positions are two out of the top three fastest growing tech occupations for 2020, with projections showing demand increasing by as much as 50% year over year.
In Michigan, you’ll find those jobs coming from major employers with data-intensive needs like GM, Ford, Whirlpool, and Dow… big name brands that have global presence and stability, and with a constant and growing need to build out their teams of accomplished data scientists.
Becoming a data scientist in Michigan and moving towards a successful career starts with the right education, work experience, and skills.
Preparing to Enroll in a Master’s Degree Program in Data Science
Preparation for a master’s degree in data science starts with an appropriate undergraduate degree, developing key proficiencies and gaining relevant work experience. While each data science master’s program will have its own admission criteria, many require at least a few years of pertinent work experience as a minimum condition for admittance. Applicants may also need to prove their competency through high scores on GRE and/or GMAT exams and fill gaps in functional knowledge through massive open online courses (MOOCs) and bridge programs.
Undergraduate Degree and Master’s Prerequisite Courses
Academically speaking, students should approach a master’s in data science with an undergraduate degree in a quantitative field, such as applied math, computer science, statistics, or engineering. The cumulative GPA should not be below 3.0. A prospective student’s undergraduate course load should be rich in key disciplines such as:
- Calculus I and II
- programming languages
- Quantitative methods
- Linear algebra
Relevant Personal and Work Experience for Admissions
Competitive data science graduate programs have the luxury of choosing students from highly compatible backgrounds. This means students who:
- Have at least five years of work experience that demonstrates technical and quantitative skills
- Have personal experience that relates to mathematics, database administration, statistics, or programming
Having a good employment record is also critical for securing the letters of recommendation that are necessary for graduate school admissions applications. Some examples of potentially qualifying work experience in Michigan may include:
- Providing IT services like database management for the University of Michigan
- Working with General Motors to provide any analytical services that relate to a range of fields from human resources or supply to productivity and sales
- Working within the healthcare industry, such as with Spectrum Health Systems, to establish, maintain, or analyze statistics
- Provide computer programming or cyber security services for the State of Michigan
Succeeding on GRE/GMAT Exams
Graduate schools expect applicants to score in at least the 85th percentile of a GRE or GMAT exam. The companies that sponsor these exams typically provide resources for preparation, or you can find other prep programs that are designed to boost your performance from third-party organizations.
GRE –The Graduate Record Exam (GRE) revised general test’s quantitative reasoning section evaluates the following:
- Arithmetic topics including integers, factorization, exponents, and roots
- Data analysis, covering topics like statistics, standard deviation, interquartile range, tables, graphs, probabilities, permutations, and Venn diagrams
- Algebraic topics such as algebraic expressions, functions, linear equations, quadratic equations, and graphing
- Geometry, including the properties of circles, triangles, quadrilaterals, polygons, and the Pythagorean theorem
Students can prepare for the quantitative reasoning section by reviewing Educational Testing Service’s (ETS) Math Review. GRE practice exams are available through the Princeton Review and Veritas Prep.
The GRE is also offered in two relevant subject tests, covering the following topics:
Physics – physics test practice book
- Classical mechanics
- Optics and waves
- Statistical mechanics
- Quantum mechanics
- Atomic physics
- Special relativity
- Lab methods and specialized topics
Mathematics – mathematics test practice book
- Introductory real analysis
- Discrete mathematics
- Probability, statistics, and numerical analysis
GMAT – The Graduate Management Admissions Test’s (GMAT) quantitative section evaluates a student’s skills as they relate to data analysis. One of the four sections of the GMAT, the quantitative section is comprised of 37 questions to be completed in 75 minutes. All of the questions pertain to data sufficiency and problem solving.
GMAT practice exams are available through the Princeton Review and Veritas Prep.
Online Data Science Bootcamps to Build Skills For Your Master’s Program or for Direct Entry into the Industry
High test scores are great, but they really don’t give you, or allow you to demonstrate, the kind of practical expertise that many data science master’s programs are looking for. One place to get both the knowledge and the hands-on skills they prefer is in a data science boot camp.
Bootcamps are laser-focused on practical applications for data science. Leaving theory by the side of the road, they dump you directly into hardcore, intensive lessons that are built around real-world datasets and solving the same sorts of challenges that employers will throw at you. It’s learning by doing, implementing a series of projects in cooperation with your close-knit student cohort and advised by instructors who are typically fresh from the front lines of data science.
Lasting only weeks or months, and costing far less than a full-fledged degree, these programs were originally the province of private organizations trying to kick-start the field by retraining math and programming professionals with the curious combination of those skills required in data science. But today you can find a wide range of bootcamps, aimed at every skill level and outcome, and offered not just by private companies but also major universities such as the MSU Data Analytics Boot Camp.
As a six-month, part-time program, available online or on campus in Detroit, this entry-level program offers unusual flexibility that is designed to work for folks who are still in school or working a regular job. But like other entry-level programs, it offers a solid base of education in data science elements like:
- Coding skills in Python and R
- AI and machine learning techniques
- Hadoop and other big data storage systems
And like other bootcamps, it comes with an array of career services that are designed to help you build your portfolio and hone your resume for immediate entry into the job market. But those same faculties can also be used to build up your CV to improve your odds of entry into a master’s program in data science.
MOOCs and Bridge Courses: Filling Gaps in Functional Knowledge
There are other, less intense ways of getting the fundamental knowledge that master’s admissions committees look for, however. One of those is something you can do on your own before you apply, while the other is offered as an option to applicants who may be good candidates but have particular deficiencies in certain core skills.
MOOCs – Massive open online courses take the form of video lectures, discussions, and interactive user forums that involve students, professors, and teaching assistants. These can be a valuable resource when it comes to filling in missing pieces in a prospective graduate student’s academic or personal history. MOOCs can be joined by anyone at any point, which gives you a lot of flexibility in how you choose to educate yourself.
Bridge Courses – Because data science incorporates skills from several fields – engineering, programming, communications, and statistics to name a few – master’s admissions might require that some students complete an academic regimen that fills gaps in their experience prior to transitioning to graduate-level coursework. Schools that house data science master’s programs make pre-master’s bridge courses available to students that have already been accepted into the program. Bridge courses typically take 15 weeks to complete, after which students would begin their master’s-level coursework.
Fundamental bridge programs:
- Analysis of algorithms
- Linear algebra
- Data structures
Bridge programs for code programming:
Earning a Master’s Degree in Data Science in Michigan
Although it’s a newly established field, undergraduate programs in data science have sprung up all across Michigan, feeding into both master’s level programs at prestigious universities like MSU and graduate certificate programs.
Students also have the option of completing their graduate degree online. Examples of online data science graduate programs available to students in Michigan include:
- Master of Science (MS) in Data Science
- Master of Information and Data Science (MIDS)
- Master of Science in Data Science (MSDS)
- Online Graduate Certificate in Data Science
- Data Mining and Applications Graduate Certificate
A full master’s degree program includes approximately 30 semester credits. Completion times for a graduate program in data science vary, depending on the options offered by the school:
- Traditional completion time – approximately 18 months or three semesters
- Accelerated completion – completion in as little as 12 months or two semesters
- Part-time – completion in as much as 32 months or five semesters
- Graduate certificate programs can be completed in one or two semesters
Core Curriculum and Immersion
Master’s students learn about core curriculum subjects that include:
- Machine learning and artificial intelligence
- Statistical sampling
- Information visualization
- Ethics and law for data science
- Data mining
- Network and data security
- Experiments and casual inference
- Applied regression and time series analysis
- Data research design and applications
Students complete an immersion process towards the end of their program, which represents a real-world application of skills developed up to that point. The immersion process involves work on a capstone project and allows students to demonstrate their competencies to their professors as well as visiting potential employers.
Key Competencies and Objectives
Students who earn a master’s degree in data science are expected to have achieved mastery of these core competencies, allowing employers to hire them with confidence that they’ll be able to hit the ground running:
- Familiarity with hash algorithms, cyphers, and secure communications protocols
- Ability to conduct association mining and cluster analysis
- Ability to run an analysis of survey data
- Ability to develop innovative design and research methods
- Ability to work in teams to achieve specific goals
- Ability to interpret and communicate results
- Ability to develop and conduct sophisticated data analyses
Career Opportunities in Michigan for Data Scientists with Advanced Degrees
Major employers like the State of Michigan, auto manufacturers, and healthcare companies like Spectrum are aggressively recruiting qualified data scientists from throughout the nation to come work here. As the state’s major employers, these exemplify the demand for data scientists that is growing across all of the state’s economic sectors, a fact that holds true even down to some of Michigan’s smallest job creators. For example, according to the Michigan Tech News, in 2019 some 37 venture-backed startups in the Detroit area were hunting for data scientists at the same time as automotive behemoth Ford Motor Company.
Many of today’s candidates with graduate degrees did not have the option of obtaining a master’s degree in data science since graduate programs specific to the field were not yet developed. That makes anyone with a freshly minted data science master’s degree intensely competitive even without a lot of on-the-job experience.
The following job listings are shown as illustrative examples only and are not meant to represent job offers or provide any assurance of employment.
Data Scientist with Ford Motor Company in Dearborn
- Work with Ford’s global data, insight, and analytics department
- Data scientist is responsible for all phases of data creation, model development, and model deployment with the goal of creating dealer profiles
- Applicants must have a master’s degree in a qualitative field like statistics, economics, quantitative finance, industrial systems and engineering, or mathematics
Data Scientist Analyst with Altair ProductDesign in Dearborn
- A global product development consultancy and subsidiary of Altair Engineering
- Responsibilities include ensuring that data is accurate, and that policies regarding quality, risk management, business management, and data management are followed
- Preferred applicants hold a master’s degree and have two years of experience in a big data environment
Data Scientist with Spectrum Health in Grand Rapids
- Work as part of Spectrum Health’s consumer and market intelligence team
- Duties involve using data to solve core business problems, as well as building and designing analytical systems to provide insight for business strategy development
- Applicants must have at least a master’s degree in a quantitative field, such as statistics, bioinformatics, epidemiology, or computer science